Goals

The project aims to enhance the performance of the Cisco proprietary algorithm used to compute network paths within the Path Computation Element (PCE) protocol, which enables multiplexing of optical paths over a shared network topology. By leveraging AI techniques, such as deep reinforcement learning and machine learning, the goal is to augment the existing PCE logic to more effectively compute feasible network paths in complex scenarios. These scenarios involve conflicting demands from limited resources, diverse service requirements, and routing constraints, which are often difficult to satisfy using the default path placement logic.